Android Scan in Developer Challenge

24 views
Skip to first unread message

Jeffrey Sharkey

unread,
May 10, 2008, 8:20:00 PM5/10/08
to zxing
Hey there ZXing community :)

I just wanted to thank you for releasing the ZXing library under an
Apache open-source license. My Android application was one of the 50
winners of the Android Developer Challenge, and ZXing was an important
part of my submission.

http://www.jsharkey.org/blog/2008/05/10/winner-of-google-developer-challenge/

In the next week or so I'll be writing up a detailed guide for getting
ZXing working with real-world barcodes on the Android emulator using
some of the methods mentioned in my documentation, such as using a
MiniDV camera through JMF.

Thanks, and I'm looking forward to meeting some of you at Google I/O.

srowen

unread,
May 10, 2008, 9:35:52 PM5/10/08
to zxing
Congratulations, that's quite an achievement. Obviously I think this
is a fantastic application of Android. We had nothing to do with the
judging of course, but would have voted for this guy if we had. 2D
barcodes are cool but 1D barcodes are far more prevalent in the real
world right now. This looks like a very polished application that goes
well beyond just decoding and provides a lot of a value to the user
with that information.

It's worth keeping in mind that 1D barcodes remain hard to scan with a
camera unless it has reasonable auto-focus and/or macro-mode
capability. With a clear shot from a webcam it's no problem; phone
cameras tend to be less capable. But improving software and improving
hardware will make that issue go away. I know Daniel has been
tinkering with basic sharpening filters in recent builds and it's
helped significantly, so I bet you will have even better results with
the latest version. I bet this will work well on the first real
Android handsets.

One other thing you can see in recent builds is that the decoder app
is continuous-scan now, instead of waiting for the user to press a
scan button. It's fast enough to do it this way even when searching
for 2D formats too; if just looking for UPC you can definitely scan
many times a second even. Hey, if you miss once or twice, no big deal,
it's just adding a fraction of a second to the response time. There is
no "miss".

I know you're releasing as GPL so we can't reuse whatever enhancements
you have made directly. It is great of course that you are releasing
it to the community, period. But if you came up with any improvements
that would benefit the core library I hope we can find a way to get
them back into this project.

On May 10, 8:20 pm, Jeffrey Sharkey <Jeffrey.Shar...@gmail.com> wrote:
> Hey there ZXing community  :)
>
> I just wanted to thank you for releasing the ZXing library under an
> Apache open-source license.  My Android application was one of the 50
> winners of the Android Developer Challenge, and ZXing was an important
> part of my submission.
>
> http://www.jsharkey.org/blog/2008/05/10/winner-of-google-developer-ch...

Brendon J. Wilson

unread,
May 15, 2008, 12:29:03 PM5/15/08
to zxing
Congrats on the AndroidScan win! However, the real-world application
you've created now raises an important question for the ZXing project:

Various methods of performing barcode recognition using a cameraphone
are patented by Neomedia and Scanbuy. They cover using a camera to
perform the recognition, as well as querying a server based on a the
recognized barcode to retrieve content, as well as alternate means to
perform recognition by taking a photo and uploading it to a server for
decoding. How is ZXing going to move forward when any real world
applications will ultimately infringe these patents? Is Google going
to license those patents? Or is the point that ZXing is simply the
implementation, and it's up to the implementer to license the patents
themselves?

Thanks for any insight.

Brendon

On May 10, 5:20 pm, Jeffrey Sharkey <Jeffrey.Shar...@gmail.com> wrote:
> Hey there ZXing community :)
>
> I just wanted to thank you for releasing the ZXing library under an
> Apache open-source license. My Android application was one of the 50
> winners of the Android Developer Challenge, and ZXing was an important
> part of my submission.
>
> http://www.jsharkey.org/blog/2008/05/10/winner-of-google-developer-ch...

srowen

unread,
May 15, 2008, 3:00:29 PM5/15/08
to zxing
I would like to be careful when writing a post about patent issues
since I am not a lawyer and am not familiar with details like these.
But I can state some facts for you; it is not clear, as you say, that
any of this is legitimately patented.

We have release original code (of course), which implements open,
public specifications. QR Codes were released in 1994 and 1D barcodes
well before then, which one should compare to the date of any patent.

I also note that the EFF has successfully lobbied the USPTO to review
some of NeoMedia's patents.

You can observe some of the repetitive debate about this issue in blog
posts like this:
http://www.gomonews.com/scanbuy/

According to this thread at least, the tussle between NeoMedia and
Scanbuy concerns only "indirect" linking, wherein a code encodes only
an opaque identifier that has meaning only when resolved against some
special server. This sort of thing is specifically not supported by
zxing, and doesn't sound to me like what Android Scan is doing.

I leave it to others to read and judge patents, and decide whether
they seem obvious or not for example. I know of no reason to believe
there are patent issues with the library that we have released. I note
also the large number of barcode readers out there, from Nokia,
3GVision, Quickmark and others that seem to have no need of licensing
arrangements, none that I have heard of.


On May 15, 12:29 pm, "Brendon J. Wilson" <brendon.wil...@gmail.com>
wrote:

Romin Irani

unread,
May 15, 2008, 11:48:25 PM5/15/08
to zx...@googlegroups.com
Hello Sean,
 
Thanks for some clarifications on the same. I am no expert on the topic but I would still like to make it more simple for everyone to understand if there are any patent issues with two use cases of how I plan to use zxing:
 
a) Decode the 2D Bar Code *using zxing* library and display the decoded text. In this case, the 2D barcode contains only text and thereby needs *NO ROUND TRIP* to the Server to decode the information.
 
b) Decode the 2D Bar Code *using zxing* library and this time the decoded text contains a url -- something like http://myserver/decode?id=12345. Now on the server side, I have an implementation that does a lookup on the id=12345 and returns some formatted text back to the mobile client.
 
Can you confirm:
1) That both (a) and (b) can be done without any patent infringement assuming that Neomedia or whoever it is still retains their patent.
2) That only (a) is OK and (b) is *NOT OK* since you could possibly say that (b) can be made sense off only via a Server?
 
Looking forward to your response.
 
Thanks
Romin.

--
-------------------------------------------------------
Romin K. Irani
romin....@gmail.com

srowen

unread,
May 16, 2008, 5:40:32 AM5/16/08
to zxing
I couldn't claim to give a definitive answer on anything related to
patents. I can just give some facts and resources.

I can say I know of no reason to believe that what this library
implements falls under any patent claims or anything of the sort. I
hope that goes without saying. (Well, Denso Wave has patented QR Code,
but, have publicly stated will not exercise this patent.) QR Code is
an ISO standard and encoding/decoding text is plainly part of that
open standard. I have not heard of any patent challenges to QR Code
itself.

I think your question is a bit broader, about what particular
applications of encoding or decoding might be patented, which is
separate from the question of whether the format, or simple act of
encoding or decoding is patented.

The discussion around the Neomedia patent that I have observed in
passing on the web, including comments from a person apparently
connected to Neomedia, suggest that they do not make a claim on
"direct" linking, which seems to mean encoding a URL, but rather
"indirect" linking. This has been repeated to me in private
correspondence, that for example Google's use of QR Codes to embed
hyperlinks in print ads is not something this person claims Neomedia
has a patent on. So, make of that what you will. I cannot say how
authoritative these interpretations are.

http://www.benmiller.com/2006/11/qrcode-url-redirect-patent-troll.html
http://theponderingprimate.blogspot.com/2007/04/neomedias-mobile-barcode-patent.html
... invite some comparisons between the date of various patents, and
the 1994 release of QR Codes.



On May 15, 11:48 pm, "Romin Irani" <romin.k.ir...@gmail.com> wrote:
> Hello Sean,
>
> Thanks for some clarifications on the same. I am no expert on the topic but
> I would still like to make it more simple for everyone to understand if
> there are any patent issues with two use cases of how I plan to use zxing:
>
> a) Decode the 2D Bar Code *using zxing* library and display the decoded
> text. In this case, the 2D barcode contains only text and thereby needs *NO
> ROUND TRIP* to the Server to decode the information.
>
> b) Decode the 2D Bar Code *using zxing* library and this time the decoded
> text contains a url -- something likehttp://myserver/decode?id=12345. Now

Dean Collins

unread,
May 16, 2008, 7:48:20 AM5/16/08
to zx...@googlegroups.com

Lol yeh right (b) part is covered by a patent my butt.

 

I have about 20 prior art examples – welcome to hubris patents, good for raising stupid money from blind investors to throw their money into a company without anyone actually looking at what it really covers.

 

Would you care to make a declaration about who you are Romin and who you work for?

 

 



Cheers,
Dean


Romin Irani

unread,
May 17, 2008, 10:08:19 AM5/17/08
to zx...@googlegroups.com
Dean,
 
Relax! I am proud to let you know that I work for InSync Software (www.insyncinfo.com) and 2D barcodes is an area that interests me because of a range of applications that could benefit from it.  
 
I asked these questions both for my knowledge and (if applicable) for other readers.
 
Fortunately - I do not practice the "Art of FUD". If you are still not sure who I am - drop me a direct line - I wish to keep this forum focussed on the topics.
 
Peace,
Romin.
 

Dean Collins

unread,
May 17, 2008, 10:14:43 AM5/17/08
to zx...@googlegroups.com

My point is very simple – there is a lot of crap in the USA market about Scanbuy and NeoMedia owning this and that etc, and none of it is court tested.

 

Build whatever you like and when they come after you simply say thanks – see you in court, they will never let it get that far as too much of their IP is tenuous at best. Just look at the crap being thrown around on Scott Schaffers blog about who owns what. http://www.theponderingprimate.com

 

Too many examples of prior art.

 



Regards,

Dean Collins
de...@cognation.net
Cognation Limited
+1-212-203-4357
+61-2-9016-4652 (Sydney indial)


Romin Irani

unread,
May 17, 2008, 10:26:05 AM5/17/08
to zx...@googlegroups.com
Dean,
 
Thanks for the information.
 
Your comments is just what I need to continue my focus on 2D BarCode enabled Applications and not worry about Scanbuy,NeoMedia,etc.
 
Romin.

 

srowen

unread,
May 17, 2008, 2:32:09 PM5/17/08
to zxing
Whoa I think we are all in agreement here. I am equally sure you and
Romin agree on these patent issues and that Romin has nothing to hide.
It's possible that after a quick look at the last message one may get
the sense he was trying to voice support for these patents by means of
a question, but, I think it really was just a question!

On May 16, 7:48 am, "Dean Collins" <D...@cognation.net> wrote:
> Lol yeh right (b) part is covered by a patent my butt.
>
> I have about 20 prior art examples - welcome to hubris patents, good for
Reply all
Reply to author
Forward
0 new messages